How to Analyze Split Test Data

Accessing Experiment Results

1. Log In: Access your Humblytics dashboard.

2. Navigate to Split Testing: Go to the "Split Testing" Tab section and select the experiment you want to analyze.

Overview Section

1. Experiment Summary: Review the experiment details including:

   • Experiment Name

   • Creation and End Dates

   • Randomization Type (Session Level or User Level)

2. Key Metrics: Examine the overall performance metrics for each variant:

   • Unique Visitors: Number of distinct users visiting each variant.

   • Page Views: Total page views for each variant.

   • Click Conversion Rate: Percentage of users who clicked on the targeted elements.

   • Form Conversion Rate: Percentage of users who submitted forms.

   • Bounce Rate: Percentage of users who left the site after viewing only one page.

   • Average Session Length: Average time spent on the site.

   • Average Scroll Depth: Average scroll depth indicating how far users scroll down the page.

Target Clicks Section

1. Click Targets: Identify the specific elements being tracked, such as buttons or links.

2. Click Data: Compare the number of clicks and conversion rates for each click target across variants:

   • URL A Clicks: Number of clicks on variant A.

   • URL B Clicks: Number of clicks on variant B.

   • Conversion Rates: Percentage of clicks that led to the desired action (e.g., form submission, purchase).

Form Submissions Section

1. Form Metrics: Review the form submission data:

   • Number of Submissions: Total form submissions for each variant.

   • Conversion Rates: Percentage of visitors who submitted forms.

2. Comparison: Compare form submission performance between variants to identify which variant led to more form submissions.

Real-Time Monitoring

1. Live Data: Monitor real-time data to track user interactions and conversions as they happen.

2. User Behavior: Observe how users interact with different variants to gain insights into their behavior.

Comparative Analysis

1. Visual Comparison: Use visual aids such as graphs and charts to compare the performance of each variant.

2. Detailed Metrics: Drill down into specific metrics to understand why one variant outperformed the other.

Determining the Winner

1. Statistical Significance: Ensure that the results are statistically significant before declaring a winner.

2. Winner Selection: Determine which variant performed better based on your predefined goals and key performance indicators (KPIs).

Implementing the Best Variant

1. Deploy the Winner: Implement the successful variant on your website.

2. Ongoing Monitoring: Continue to monitor the performance of the implemented variant to ensure sustained improvements.

Next Steps

1. Iteration: Use the insights gained from the analysis to design new experiments and continue optimizing your website.

2. Documentation: Document the results and key learnings from each experiment for future reference.

By following these steps, you can effectively analyze split test data, make data-driven decisions, and continuously improve your website's performance.